imapsync very poor performance

[Franz Skale]

Start dovecot in the foreground using option -F
enable full debugging. (Dovecot.conf).
Or start dovecot unsing strace.
These are standard debugging options you can try.
ALso check your DNS setup. (resolv.conf)
Do you use dovecot indexes ?


Rgds.

Franz


Am 02.12.13 19:27, schrieb Christian Hügel:
> This has to do with Centos 6.5. There is no issue with Centos 6.4; I've
> just setup up a VM. The question is which files causes this issue.
>
> Christian
>
> Am 02.12.2013 17:07, schrieb Franz Skale:
>>
>>
>> Check for hanging threads.
>> Also check ,that your sysctl.conf maxfilenr is big enough.
>> Change your init script accordingly. (ulimit -n 4096 before the start
>> option)
>>
>> Check sysctl.conf:
>> fs.inotify.max_user_instances = 4096
>> fs.file-max = 2000000
>>
>> do: sysctl -p after changing the sysctl.conf
>>
>> Check the dovecot error log. /var/log/dovecot-error.log (on debian systems)
>>
>>
>> Rgds
>>
>>
>> Franz
>>
>>
>>
>>
>> Am 02.12.13 17:01, schrieb Christian Hügel:
>>> Hi Franz,
>>>
>>> with those options dovecot specially.
>>>
>>>> login_max_processes_count = 1024
>>>> login_max_connections = 2048
>>> the dovecot server hangs and isn´t accessible anymore.
>>>
>>>
>>> Christian
>>>
>>> Am 02-12-2013 16:47, schrieb Franz Skale:
>>>> Check your dovecot settings.
>>>> Also, the the dovecot error log.
>>>> Typical settings of a busy host:
>>>>
>>>> mail_max_userip_connections = 100
>>>> login_max_processes_count = 1024
>>>> login_max_connections = 2048
>>>> In the init script use ulimit -n 4096
>>>>
>>>> Then try again.
>>>>
>>>> Rgds.
>>>>
>>>> Franz
>>>>
>>>>
>>>>
>>>> Am 02.12.13 15:41, schrieb Christian Hügel:
>>>>> i,
>>>>>
>>>>> I´m running imapsync-1.569-1.el6.noarch on a Centos 6.5 x64 system
>>>>> with 8GB RAM and a dual 3Ghz AMD cpu. The sync runs on the "target
>>>>> server and the source server is a dovecot 1.2.1
>>>>> Here are some stats:
>>>>>
>>>>> 0.01 msgs/s 63.780 KiB/s ETA: Mon Dec 2 22:31:21 2013 28528 s 328 msgs
>>>>> left
>>>>> 0.01 msgs/s 65.135 KiB/s ETA: Tue Dec 3 01:17:36 2013 38356 s 327 msgs
>>>>> left
>>>>> 0.01 msgs/s 65.076 KiB/s ETA: Mon Dec 2 21:44:20 2013 25560 s 326 msgs
>>>>> left
>>>>> 0.01 msgs/s 65.238 KiB/s ETA: Mon Dec 2 23:08:44 2013 30484 s 325 msgs
>>>>> left
>>>>> 0.01 msgs/s 65.181 KiB/s ETA: Mon Dec 2 21:26:25 2013 24344 s 324 msgs
>>>>> left
>>>>> 0.02 msgs/s 65.122 KiB/s ETA: Mon Dec 2 20:18:12 2013 20251 s 323 msgs
>>>>> left
>>>>> 0.02 msgs/s 65.235 KiB/s ETA: Mon Dec 2 19:46:36 2013 18332 s 322 msgs
>>>>> left
>>>>>
>>>>> and the cpu is during the sync at 100% The size of the mails are
>>>>> normal no attachments so I realy don´t know where´s the issue here.
>>>>> The size of the mailbox is only 300MB but I have a couple of huge
>>>>> mailboxes (>17GB). With this speed it´s quite impossible to migrate my
>>>>> mailboxes to the new kolab 3.1 server
>>>>>
>>>>> Can you please help me out with this issue? Thank you
>>>>>
>>>>> Christian
>>> _______________________________________________
>>> users mailing list
>>> users at lists.kolab.org
>>> https://lists.kolab.org/mailman/listinfo/users
>>
>>
>>
>> _______________________________________________
>> users mailing list
>> users at lists.kolab.org
>> https://lists.kolab.org/mailman/listinfo/users
>>
>
>
> _______________________________________________
> users mailing list
> users at lists.kolab.org
> https://lists.kolab.org/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.kolab.org/pipermail/users/attachments/20131202/739a06d9/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4495 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.kolab.org/pipermail/users/attachments/20131202/739a06d9/attachment.p7s>